Balasubramanian Thiagarajan

About Me

I'm a data informatics and machine learning engineer with experience building and scaling AI‑driven systems. My background spans retrieval‑augmented generation applications, recommender systems, and data pipelines. I enjoy experimenting with LLMs and leveraging data science to solve real‑world problems.

Education

Master of Science in Data Informatics

University of Southern California, Los Angeles, CA, USA — Jan 2018 – Dec 2019, GPA: 3.87 / 4

Introduction to Data Management
Machine Learning for Data Informatics
Foundations and Applications of Data Mining
Information Visualization
Applied Natural Language Processing
Building Knowledge Graphs
Computer Vision

Bachelor of Engineering in Computer Science and Engineering

Anna University, Chennai, India — Jun 2013 – Apr 2017, GPA: 8.16 / 10

Design and Analysis of Algorithms
Database Management Systems
Software Engineering
Artificial Intelligence

Experience

Intermediate Software Developer – AI — Global Relay

Vancouver, BC | Apr 2025 – Present

Contributing to maintaining a Retrieval‑Augmented Generation (RAG) search application using Milvus and Large Language Models for risk detection in enterprise communications.
Scaled system architecture to ensure high performance and reliability for large‑scale message surveillance.
Conducted model tuning experiments; improved retrieval accuracy by 25% using a custom trained BM25 model.

Machine Learning Engineer (P3) — Zillow (Canada) Inc.

Vancouver, BC | Jan 2023 – Dec 2024

Built and maintained end‑to‑end ML systems on AWS, Spark, and Airflow to deliver home recommendations to 20M+ users across Zillow and Trulia platforms.
Trained and evaluated models and performed A/B experimentation, yielding $1M+ estimated gains annually.
Developed streaming applications using Kafka, Metaflow and Kubeflow for near real‑time home recommendations.
Led migration of our A/B experimentation platform from an internal tool to Split.io.

Machine Learning Engineer (P2) — Zillow Inc.

Seattle, WA | Feb 2020 – Dec 2022

Contributed to building and scaling ML pipelines for home recommendation systems on Zillow and Trulia platforms.
Implemented online A/B experiments to optimize user engagement and recommendation quality.

Publications

Volume Governing for Email and Push Messages

ACM RecSys '22 Poster

Developed notification governance batch pipeline hosting ML models for volume control in Zillow for ~40M users.

Modeling engagement in long‑term, in‑home socially assistive robot interventions for children with autism spectrum disorders

Science Robotics

Analyzed interactions and implemented ML algorithms to model user engagement in socially assistive robotics.

A Machine Learning Approach for Prediction of On‑Time Performance of Flights

IEEE/AIAA 36th Digital Avionics Systems Conference (DASC) 2017

Led development of a two‑stage predictive model for flight on‑time performance and built a real‑time decision support tool.